[Paper Review] PV Integration in Low-Voltage Feeders with Demand Response

Xiangkun Li, Theodor Borsche+1 more|arXiv (Cornell University)|Aug 7, 2014

Photovoltaic System Optimization Techniques7 references3 citations

TL;DR

This paper proposes a demand response (DR) control strategy using electric water heaters (EWHs) to increase photovoltaic (PV) hosting capacity in low-voltage (LV) distribution feeders. By leveraging only real-time power measurements at the substation and employing model predictive control (MPC) to optimize EWH scheduling, the method reduces voltage violations and PV curtailment, achieving over 50% increased PV hosting capacity with fewer than two load activations per day on average.

ABSTRACT

Increased distributed Photo-Voltaic (PV) generation leads to an increase in voltages and unwarranted backflows into the grid. This paper investigates Demand Response (DR) with Electric Water Heaters (EWHs) as a way to increase the PV hosting capacity of a low-voltage feeder. A control strategy relying only on power measurements at the transformer is proposed. Flexible loads are optimally dispatched considering energy acquisition costs, a PV shedding penalty, and power and energy constraints. Furthermore, grouping of loads and PV plants is investigated, and switching penalties are used to reduce the unnecessary switching of loads. It is shown that this strategy can substantially increase the PV hosting capacity of a Low-Voltage (LV) feeder, even when only basic controllability is available.

Motivation & Objective

• To address voltage rise and reverse power flow issues caused by high PV penetration in LV distribution networks.
• To investigate how demand response via controllable loads like EWHs can increase PV hosting capacity without requiring advanced metering or real-time communication at every node.
• To evaluate the trade-off between control granularity, switching frequency, and PV curtailment under realistic constraints such as grouped load control and forecast uncertainty.
• To assess the feasibility of using only substation-level power measurements and AMI data for optimal dispatch of flexible loads.
• To determine whether limited controllability and switching penalties can still yield significant improvements in PV integration.

Proposed method

• A model predictive control (MPC) framework is used to optimize EWH charging schedules based on real-time power flow measurements at the LV transformer.
• The optimization minimizes energy acquisition costs, PV shedding penalties, and respects active and reactive power limits, as well as energy capacity constraints of EWHs.
• Load and PV units are grouped into clusters to reflect real-world limitations of legacy control systems like ripple-control, reducing control complexity.
• A switching penalty is introduced in the cost function to limit the number of load state changes, reducing wear and improving customer acceptance.
• The approach uses historical and forecasted PV generation and load data from Swiss households, with uncertainty handled via receding horizon control and stochastic modeling.
• The control strategy is validated on a realistic 400V LV distribution network with two feeders and 20 nodes, using actual PV and load profiles.

Experimental results

Research questions

• RQ1 Can demand response using EWHs significantly increase the PV hosting capacity of a low-voltage distribution feeder with only substation-level power measurements?
• RQ2 How does grouping of flexible loads affect the performance and required switching frequency of the DR strategy?
• RQ3 To what extent can switching penalties reduce unnecessary load switching while maintaining high PV hosting capacity?
• RQ4 How does forecast uncertainty impact the performance of the control strategy, and what control techniques are most effective in handling it?
• RQ5 What is the trade-off between PV curtailment, energy cost, and load switching frequency under realistic system constraints?

Key findings

• The proposed DR strategy increased the PV hosting capacity of the LV feeder by over 50% compared to a baseline without control.
• The method achieved this improvement while limiting EWH activation to fewer than two times per day on average, significantly reducing switching-related wear.
• Even with grouped load control and limited controllability, the system maintained high performance, indicating that full individual control is not necessary.
• The inclusion of a switching penalty in the optimization reduced unnecessary load switching with minimal impact on overall system performance.
• The strategy effectively managed voltage rise, keeping it within the 3% limit specified by the DACHCZ code, even under high PV penetration.
• The results show that using only transformer-level power measurements and AMI data for daily load energy profiles is sufficient for effective control, reducing the need for advanced metering at every customer.